pmariglia/foul-play

A Pokemon Showdown Battle AI

What it solves

Foul Play is a bot designed to automate Pokémon battles on the Pokemon Showdown platform. It allows users to run a bot that can compete in single battles across all generations of the game, though it currently does not support Dynamax or Z-moves.

How it works

The bot integrates with the Pokemon Showdown websocket URI to communicate with the game server. It utilizes a specialized search engine called poke-engine (built in Rust) to analyze battle states and search through possible moves to determine the optimal play.

Who it’s for

Players of Pokemon Showdown who want to automate their battles or developers interested in game AI and search-based decision making in turn-based strategy games.

Highlights

  • Multi-generational support: Capable of playing single battles across all Pokémon generations.
  • Search-based AI: Uses a dedicated engine to calculate moves rather than simple scripts.
  • Docker support: Provides a Docker image for easier deployment and environment setup.
  • Flexible configuration: Configurable via command-line arguments for usernames, passwords, and battle formats (e.g., gen9randombattle).

Related

  • Project
    hsahovic/poke-env

    A Python library for building scripted agents and reinforcement learning workflows for Pokemon Showdown battles.

  • Project
    lihongxun945/gobang

    A browser-based Gobang AI implemented using the Minimax algorithm and Alpha-Beta pruning, featuring a comprehensive evaluation system for measuring AI strength.

  • Project
    Pbatch/ClashRoyaleBuildABot

    An automation bot for Clash Royale that uses an advanced state generator and object detection for educational and research purposes.

  • Project
    Farama-Foundation/ViZDoom

    A Doom-based AI research platform for developing and testing bots that play the game using only visual information, primarily for reinforcement learning research.

  • Project
    pasky/pachi

    A modular framework and game engine for Go (Weiqi/Baduk) that combines Monte Carlo Tree Search with deep learning to provide a high-strength opponent and analysis tool.